The general purpose of this research effort is to better understand how these new policy initiatives affect important farm and farm household decisions, and in turn affect the supply of agricultural products and the farm sector`s contribution to environmental quality, the supply of valued non-commodity outputs, and farm household well-being. Our working hypothesis is that decisions to participate in programs to promote environmental or other “multifunctional” goals of agriculture are inextricably tied to the life-cycle of the farmer and to decisions about cropping patterns livestock production, work off the farm, and effective risk management. To accomplish this understanding, we will model these important decisions jointly using advanced econometric techniques. Having modeled these joint decisions, it will be possible to use the models initially to simulate the effects of future changes in policy at the farm household level particularly with respect to farm production, farm family incomes and wealth, livelihood strategies, and participation in environmentally related programs. To the extent that new policy initiatives substantially alter farm-household behavior, this simulation capacity at the micro level is critical to predicting major changes in the farming and the broader agricultural sector at both regional and national levels.

The gradual redirection of U.S. farm policy away from traditional commodity programs is in part due to the need to reduce levels of domestic support required by the General Agreements on Tariffs and Trade (GATT) and the World Trade Organization (WTO). However, there is also growing recognition that farm policy must recognize the increasing diversity of the agricultural sector and must expand programs that address the environmental concerns related to agricultural production and promote directly the production of non-commodity outputs from agriculture that are highly valued by society. As this new policy agenda unfolds, some of its effects on the well- being of farm households will be revealed in national income accounts; the effects of the provision of these important non-commodity outputs will be reflected in economy-wide indicators of natural resource use, etc. But these aggregate data will mask both the processes by which these changes come about and the effectiveness of various policy initiatives.

Some additional analysis of the estimated technical efficiencies for farm production and farm household production was undertaken to better understand the extent to which these distributions of technical efficiencies differ depending on whether the farm participates in the Conservation Reserve Program (CRP), off-farm work, or both. These distributions of technical efficiencies for the four sub-groups of farmers (those participating only in CRP, those participating only in off-farm work, those participating in both, and those participating in neither) are ranked by stochastic dominance criteria. These results were incorporated into a paper that has been submitted for publication in an academic journal. Additional analysis of the estimated distributions of farm household income (accounting for participation in CRP and for the farm operator and spouse working off the farm) was undertaken. The additional analysis was to determine the extent to which the mean and variability of farm household income for household in the nine mutually exclusive sub-groups of households participating in various combinations of the activities differed.

Our measures of technical efficiency suggest while removing both land and labor from agricultural production is the most effective strategy to reallocate resources between the farm and the household, removing labor alone through off-farm work still improves resource allocation, but this is not true when only land is taken out of production. Our empirical results also suggest that any change in policy, etc., that affects the probability of participation in CRP or off-farm work affects mean household incomes and the variability of income across farms. The probability of participation in CRP accounts for from 31 percent of estimated farm household income for farms participating only in CRP to 4 percent for the group where only the operator works off farm. The standard deviation in income falls by between 22 percent and 27 percent among farms participating in CRP. The policy implications of these new empirical results were discussed informally with economists and administrators within the Economic Research Service (ERS), USDA. Because of the current debate over the appropriate statistical methods for estimating standard errors of coefficients in economic models to account for the complex sampling involved in the Agricultural Resource Management Survey data, the principal investigator was asked to review the report of the panel to review USDA`s Agricultural Resource Management Survey that was established by the National Research Council. This report is entitled Understanding American Agriculture, Challenges for the Agricultural Resource Management Survey.

Economic Research Service, USDA